196 PROCEEDINGS OF SECTION C. 



4.— ALKALINE ROCKS OF THE COOK AND SOCIETY ISLANDS. 

 By PROFESSOR P. MARSHALL, M.A., D.Sc, F.G.S Otago University, Dun., NZ. 



[Plates xvii -xix.] 



As a member of the Committee appointed to investigate the alkaline 

 rocks of Australasia it was arranged that I should as far as possible 

 study the distribution of these rocks in the Pacific. I was able 

 to visit Raratonga and Mangaia Islands in the Cook Group, and 

 Tahiti, Raiatea and Huaheine in the Society Group. At Raratonga 

 I stayed for six days — ^February 15th to February 22nd, 1910 ; at 

 Tahiti part of two days — February 25th and 26th ; at Raiatea part 

 of two days — February 24th and 28th ; at Huaheine, a few hours 

 on February 27th ; at Mangaia, a few hours on March 3rd. This 

 time, of course, did not allow of a complete inspection of any of 

 the islands, though Raratonga and Mangaia were traversed more 

 widely than the others. The following notes will give an idea of 

 the distribution of alkaline rocks in these islands : — 



Tahiti. — ^The time at my disposal did not allow of a visit to the 

 Papenoo Valley, and I can, therefore, add nothing to the descrip- 

 tions of plutonic rocks given by Lacroix. Frere Alain, Principal of 

 the Brothers School at Tahiti, kindly gave me specimens of rocks 

 that he had gathered there. These specimens are rolled pebbles. 

 They include nepheline syenite, mica syenite, tinguaite and thera- 

 lite, which agree satisfactorily with the specimens described by 

 Lacroix. In one specimen the hornblende is almost entirely 

 replaced by irregular plates of biotite, while the pinkish augite is 

 surrounded with a green mantle ; the sphene is surrounded by a 

 black border of magnetite. 



Lacroix has lately described a hauynophyre from the Papenoo 

 Valley and from Vairao. I found a similar rock in great quantity 

 at Arue, six miles south-east of Papiete along the coast. In general 

 it answers well to the description of the Papenoo specimen, for not 

 only is the hauyne in moderately large crystals, but in small ones 

 as well, and black inclusions are crowded on the margin. There 

 is also this difference, hornblende is present in considerable quan- 

 tity in my specimens, though in those of Lacroix " mais I'amphibole 

 y manque toujours." There is also a small quantity of felspar, but 

 I cannot identify the species. Lacroix observes that the glass, 

 which is very abundant, is heterogeneous, varying from brown to 

 colourless. In all my specimens it is colourless, and the question 

 arises whether such an abundance of colourless glass could form the 

 unindividualised residue of such a rock. To test the nature of this 

 material some of the powdered rock was warmed with dilute nitric acid. 

 It was found that 68-90 per cent, of the rock passed into solution. 

 The dissolved matter contained SiOgSS-ST, A120316-48, Fe2033-12, 

 CaO 1-47, Ti02-56, alkalies 10-20, H20"-3-20. 



When allowance is made for the hauyne and iron oxide dis- 

 solved, the remaining matter has the probable composition Si02 

 32-57, AI2O3 15-88 ; alkalies 9-30, H2O 3-20, or a percentage compo- 

 sition of SiOs 53-1, AI2O3 26-9; alkalies 15-0, H2O 5-3. This 



